Digital transition in dairy production:(SSA)

Description

Research Hypothesis Digital technology adoption in SSA dairy systems significantly improves productivity, operational efficiency, and market integration, but adoption barriers (e.g., infrastructure, gender disparities) vary across countries due to contextual factors. Data Collection Methodology Aspect Details Scope 137 dairy farms across Ghana (39), Kenya (41), Nigeria (16), Tanzania (16), Uganda (25) Timing November–December 2023 Tools Closed-ended smartphone questionnaires (translated to local languages) Sampling Hybrid: Random (representativeness) + Purposive (targeting tech-adopting farms) Variables Digital tool usage, barriers (cost/infrastructure), efficiency gains, gender roles, market impacts Ethics University-approved protocols; informed consent obtained Key Findings 1. Efficiency Gains: o Ghana showed highest improvement (mean: 9.38/21 codes) due to mobile/IoT tools. o Nigeria/Uganda lagged (means: 3.75–3.36). 2. Adoption Barriers: o Highest in Nigeria (mean: 20.25/33 codes: cost/infrastructure). o Lowest in Uganda (mean: 5.20). 3. Marketing Impact: o Tanzania/Ghana led in value-chain digitization (e.g., e-commerce platforms). 4. Gender Disparities: o Severe in Ghana (12:1 male-farmer ratio) vs. parity in Kenya/Tanzania. Data Interpretation Dataset How to Interpret Use Cases Q10 (Efficiency) Higher codes = advanced impacts (e.g., Code 15: AI-driven decisions) Prioritize high-impact tools (e.g., IoT over SMS) Q13 (Barriers) Codes 1–10 = structural (cost); 11–20 = technical (skills); 21–33 = social (gender) Target interventions (e.g., Ghana: subsidize costs) LDA Clustering (Fig 1) Uganda/Ghana = distinct clusters → country-specific adoption patterns Customize policies per country Gender Data (Q19) Female participation correlates with tech adoption (r=0.68, p<0.05) Design women-focused digital literacy programs Notable Conclusions • Tailored Solutions Needed: o Ghana: Address cost barriers despite efficiency gains. o Nigeria: Invest in electricity/internet. o Uganda: Scale low-barrier model regionally. • Gender Inclusion: Training + finance access for women farmers boosts adoption. • Stepwise Tech Integration: Start mobile-based (SMS/market apps), then scale to AI/IoT. Data Reusability • Access: Restricted due to confidentiality; contact author (c.vuvor@studenti.uniss.it) for requests. • Supplementary Files: S1 (questionnaires), S2 (qual-quant methodology), S3 (coding schemes) enable methodological replication. • Aggregated Data: Tables 2–4 support cross-country benchmarking (e.g., barrier severity indices). Key Insight: Digital tools can close SSA’s dairy supply-demand gap if deployed contextually—addressing gender, infrastructure, and cost barriers

Steps to reproduce

1. Sampling Design • Target Regions: Dairy hubs in Ghana (Northern/Greater Accra), Kenya (Kiambu), Nigeria (Jos), Tanzania (Arusha), Uganda (Mbarara). • Selection: o Random sampling: 70% of farms (96/137) for representativeness. o Purposive sampling: 30% (41/137) tech-adopting farms via local coordinators. 2. Data Collection Protocol • Tool: Closed-ended questionnaire (Supplementary File S1) administered via google form on Android smartphones. • Translation: Professionally translated to local languages (Swahili, Twi, etc.) + back-translation validation. • Fieldwork: o 23 trained enumerators (5/country). o Nov–Dec 2023; 45 min/farm average. o Variables:  Digital tools used (mobile/IoT/AI).  Barrier severity (5-point Likert scales).  Milk yield/income changes (numerical).  Gender roles (categorical: male/female). 3. Qual-to-Quant Transformation • Coding Framework (Supplementary S2): o Unitization: Segmenting text responses. o Categorization: Thematic grouping (e.g., "high costs" → Barrier Code 5). o Validation: Intercoder reliability >85% (Cohen’s κ=0.79). • Output: Numerical matrices (e.g., Q13 barriers → 33 binary variables). 4. Analytical Workflow Code, Download Raw Data, Data Cleaning in R, Qual-to-Quant Coding Statistical Analysis (LDA Clustering) • Software: R v4.4.1 + packages: o MASS for LDA. • Outputs: o Country clusters (Fig 1). 5. Reproducibility Materials • Provided: o Questionnaires (S1). o Coding schemes (S2-S3). • Constraints: o Raw data restricted (confidentiality). o Farm coordinates anonymized. 6. Ethical Compliance • University of Sassari approval (Ref: #AGR-2023-11). • Digital consent via google form. Key for Reproduction: 1. Recruit local coordinators for farm access. 2. Use Supplementary S1-S3 for instrument/coding consistency.

Institutions

• Universita degli Studi di Sassari Dipartimento di Agraria

Categories

Funders

• PNRR

Licence